Most Genetic Epidemiology Branch investigations evaluate the contributions of host susceptibility and environmental exposure in the development of cancer. In family studies, the host susceptibility measure is frequently an alteration in specific gene(s). These studies tend to be very long term with varying activity. Although two genes associated with melanoma susceptibility have been identified (CDKN2A and CDK4), alterations in these genes are found in only a small percentage of melanoma-prone families. The search for other genes continues;in collaboration with an international consortium (GenoMEL), a search for a new melanoma susceptibility genes continues both within families and a genome-wide association study. In the American melanoma-prone families, we evaluated 233 tagging SNPs in 21 genes at 9p21 among 562 individuals (183 melanoma pts) from 53 families (23 CDKN2A+ and 30 CDKN2A -) to try to identify additional genes. Single SNP- and gene-based regressions analyses were used to assess the risk of melanoma, nevus count, skin complexion, and tanning ability associated with the SNPs and genes. SNP rs7023329 in MTAP was associated with number of nevi (P(trend) = 0.003), confirming a recent finding by a genome-wide association study in genetically enriched melanoma cases and controls. In addition, 3 SNPs in ACO1 showed the strongest association with melanoma risk (rs7855483 [P(trend) = 0.002];rs17288067 [P (trend) =0.0009];rs10813813 [p (trend) = 0.005]). Gene-based analyses also demonstrated that ACO1 was significantly associated with melanoma (p=0.0004). These results are consistent with other genes in 9p21 modifying melanoma risk. We continue to accrue and evaluate new families in both the U.S and Italy. We have continued to evaluate families of individuals with heritable retinoblastoma and melanoma. The study of familial chordoma, a rare, low-grade, malignant bone tumor derived from remnants of the notochord, was expanded to include additional families. With linkage analyses showing evidence of linkage in a different chromosomal area, we extensively investigated genes in the region. Using high resolution array-CGH, we identified unique duplications of a region on 6q27 in four multiplex families. This region only contains the the T (brachyury) gene, which is important in notocord development, and is expressed in most sporadic chordomas. This is the first example of gene duplication conferring major familial susceptibility to cancer. Studying families with lymphoproliferative cancers has been a long-standing interest. We described a potential precursor lesion for CLL, monoclonal B-cell lymphocytosis (MBL)several years ago. In the general population, MBL increases with age with a prevalence of 5-9% in individuals over age 60. We conducted multi-parameter flow cytometry among 505 first-degree relatives with no personal history of lymphoproliferative disorders from 140 multiplex families with CLL. Seventeen percent of relatives had MBL. Age was the most important determinant;the probablity of developing MBL by age 90 was 61%. MBL clustered in specific families, but this clustering was independent of the number of known cases of CLL. Males had a higher risk of MBL (p = 0.04), similar to the male predominance of CLL. MBL patients had higher lymphocyte counts, and higher B-cell counts than the normal relatives. We also continued a family study of Xeroderma pigmentosum in collaboration with CCR investigators to assess risk of cancer in XP heterozygotes. Data collection is underway. We have also documented that mothers carrying affected children with tricothiodystrophy have more pregnancy complications than when carrying unaffected children.